Research map / Large language models / Language, safety and interpretability
Synchronous Speech Recognition and Speech-to-Text Translation with Interactive Decoding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-De | BLEU: 19.08 -> 19.82 | Helped |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Fr | BLEU: 30.08 -> 29.79 | Hurt |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Zh | BLEU: 20.99 -> 21.68 | Helped |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Ja | BLEU: 20.87 -> 19.60 | Hurt |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-De | WER: 15.18 -> 14.76 | Helped |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Fr | WER: 13.04 -> 12.87 | Helped |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Zh | WER: 13.43 -> 13.38 | Helped |
| use interactive attention to exchange ASR and translation decoder information | Interactive · TED En-Ja | WER: 14.01 -> 13.91 | Helped |
| increase interactive attention weight from λ=0.0 to λ=0.3 (ablation) | Interactive · TED En-Zh test | BLEU: 20.60 -> 21.01 | Helped |
| increase interactive attention weight from λ=0.0 to λ=0.3 (ablation) | Interactive · TED En-Zh test | WER: 13.43 -> 13.24 | Helped |
| increase interactive attention weight from λ=0.3 to λ=1.0 (ablation) | Interactive · TED En-Zh test | BLEU: 21.01 -> 20.13 | Hurt |
| increase interactive attention weight from λ=0.3 to λ=1.0 (ablation) | Interactive · TED En-Zh test | WER: 13.24 -> 14.52 | Hurt |
Closest papers by meaning
- High-Fidelity Simultaneous Speech-To-Speech Translation (ICML 2025 · similarity 0.89)
- TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation (NeurIPS 2024 · similarity 0.87)
- Divergence-Guided Simultaneous Speech Translation (AAAI 2024 · similarity 0.86)
- Scalable Multilingual Multimodal Machine Translation with Speech-Text Fusion (ICLR 2026 · similarity 0.86)
- LAMA-UT: Language Agnostic Multilingual ASR Through Orthography Unification and Language-Specific Transliteration (AAAI 2025 · similarity 0.85)
- Speech Recognition Meets Large Language Model: Benchmarking, Models, and Exploration (AAAI 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.